A Study of the Regulatory Input Parameters in the Granularity Adjustment of the Gordy-Lütkebohmert Model and a New Method to Calibrate Them to Rating Grades
21 Pages Posted: 1 Mar 2019 Last revised: 2 Dec 2021
Date Written: February 19, 2019
Abstract
The Gordy-Lütkebohmert model marks the golden standard for determining the granularity adjustment in regulatory capital assessments. The granularity adjustment is necessary for an approximation of the effect of undiversified idiosyncratic risk in not infinite fine grained portfolios. Internal ratings-based (IRB) risk weights of Basel II and III assume that idiosyncratic risk is diversified away on portfolio level, therefore implicitly assuming that the number of individual exposures in the portfolio trends towards ∞.
The Gordy-Lütkebohmert model is used in practise by many regulatory driven approaches as it develops a parsimonious, analytical traceable and easy to implement formula for determining the granularity adjustment. It requires certain input parameters, which are
a) dependent from underlying exposures of the portfolio (bank inputs)
b) and a set of external (regulatory) parameters.
This paper examines the regulatory input parameters and develops a new parsimonious method to calibrate and map them to rating grades.
The new method become especially important for banks with less diversified but high rated loan portfolios (e.g. public sector banks) where Granularity Adjustments are a significant part of the overall capital requirement and the correction factor due to the in this paper described calibration method is large.
Keywords: Gordy-Lütkebohmert Model, Granularity Adjustment, Regulatory Capital Assessment, IRB Risk Weights, Model Calibration, Portfolio Credit Risk
JEL Classification: C36, G28
Suggested Citation: Suggested Citation